THE NEUROSCIENCES AND MUSIC II I—DISORDERS AND PLASTICITY

Impaired Memory for Pitchin Congenital Amusia

Nathalie Gosselin, Pierre Jolicœur, and Isabelle Peretz

International Laboratory for Brain Music and Sound Research, Department ofPsychology, Universite de Montreal, Montreal, Quebec, Canada

We examined memory for pitch in congenital amusia in two tasks. In one task, we variedthe pitch distance between the target and comparison tone from 4 to 9 semitones andinserted either a silence or 6 interpolated tones between the tones to be compared. Ina second task, we manipulated the number of pitches to be retained in sequences oflength 1, 3, or 5. Amusics’ sensitivity to pitch distance was exacerbated by the presenceof interpolated tones, and amusics’ performance was more strongly affected by thenumber of pitches to maintain in memory than controls. A pitch perception deficitcould not account for the pitch memory deficit of amusics.

Key words: short-term memory; pitch; music; congenital amusia

Introduction

Congenital amusia is characterized by adeficit in musical pitch perception that cannotbe explained by hearing loss, brain damage,intellectual deficiencies, or lack of exposure.1

For example, amusic individuals have difficultyin recognizing familiar melodies and detect-ing out-of-key notes inserted in melodies. Be-havioral and electrophysiological studies haveshown that the core deficit underlying congen-ital amusia is a deficit in the fine-grained dis-crimination of pitch.2–4 More specifically, amu-sic subjects have problems in detecting a pitchdeviation of less than one semitone in a repeat-ing tone sequence.3

Deficits in fine-grained pitch perception maybe associated with pitch memory problems. Forexample, Foxton and collaborators2 found adeficit in the discrimination of two sequences,even when these were made of pitch intervalsthat amusics were able to discriminate. Sim-ilarly, Griffiths and collaborators,5 found that

Address for correspondence: Isabelle Peretz, BRAMS-Pavillon 1420Mont-Royal, University of Montreal, Montreal, Quebec, CanadaH2V 4P3. Voice: +1514 343 5840; fax: +1514 343 2175.isabelle.peretz@umontreal.ca

amusic individuals were worse than normalsin a task that required discrimination of twotones separated by a variable time interval. Thepitch interval between the two tones exceededdiscrimination thresholds, so that poor discrim-ination could not easily account for this mem-ory impairment.

The goal of the present study was to assessfurther the possibility that individuals suffer-ing from congenital amusia have pitch mem-ory deficits. Instead of examining the influ-ence of time over retention, we focused hereon interference and sequence length effectson pitch retention accuracy. Nine amusics andnine matched controls were tested in two pitchmemory tasks, both using a same–different clas-sification task. In a single-tone comparison task,we tested the impact of the presence of interpo-lated tones (i.e., interference) between the targetand comparison tone as well as the influenceof pitch distance between the tones to be com-pared. In a pitch sequence comparison task, wetested the influence of sequence length on pitchmemory by presenting one, three, or five tonesin the target and comparison sequences. Wepredicted that performance by amusic personswould be abnormally affected by both interfer-ence and sequence length.

The Neurosciences and Music III—Disorders and Plasticity: Ann. N.Y. Acad. Sci. 1169: 270–272 (2009).doi: 10.1111/j.1749-6632.2009.04762.x c© 2009 New York Academy of Sciences.

270

Gosselin et al.: Impaired Memory for Pitch in Amusia 271

Method

Participants

Nine amusics and nine controls were testedin the single-tone comparison task and sevenfrom each group also participated in the tone-sequence comparison task. Amusic participantswere matched to controls in terms of age, gen-der, education, and musical training. All amusicindividuals had normal intellectual, memory,and language abilities. In contrast, they per-formed 2 standard deviations below the meanscore obtained by normal controls on the Mon-treal Battery of Evaluation of Amusia.6

Material and Procedure

Participants were tested with two pitch mem-ory tasks in a counterbalanced order across par-ticipants. In the single-tone comparison task,they were required to compare two single tonesseparated by a retention interval and decidewhether they were the same or different. Halfthe pairs were made of identical tones. Whendifferent, the pitch distance between the twotones varied from 1 to 3 tones. The inter-tone(retention) interval was either empty (no inter-fering tones, block 1) in 24 trials, which we callthe silence condition, and was presented first;or it was filled with 6 distractor tones, whichwe call the interference condition, in 96 tri-als (block 2). The duration of the inter-toneretention interval was 1650 ms and the du-

ration of the target and comparison tone was300 ms.

In the pitch sequence task, two sequencesof tones varying in pitch were presented, sep-arated by a 2-s silent retention interval. Thetask was to decide whether the two sequenceswere the same or different. Half the pairs con-tained identical sequences. The length of thesequences varied from trial to trial with 1, 3,or 5 tones, in 180 trials (60 for each sequencelength). When different, the sequences differedin the permutation of two adjacent tones (forsequences of length 3 or 5a), or in the presenta-tion of a different tone (for sequences of length1). The tones were randomly selected among 6diatonic tones in the same octave.

Results

In the silence condition of the single-tonecomparison task, amusics were impaired whenthe pitch difference was 1 tone (t16 = –2.075;P < 0.05), and their performance at 1 toneof pitch distance was barely significantly abovechance (t8 = 2.31; P = 0.05). In contrast, amu-sics did not differ from controls when pitchdistance was 2 or 3 tones (t16 = –0.956 andt16 = –0.905, respectively; both P > 0.05).To explore the effect of interfering tones on

a The tone change in the pitch sequence task can introduce a contourchange.

Figure 1. Mean percentages of hits minus false alarms obtained in the single-tone comparison task asfunction of pitch distance and conditions (silence and interference) for amusics (A) and controls (B). Verticalbar indicates standard error.

272 Annals of the New York Academy of Sciences

Figure 2. Mean percentage of hits minus falsealarms obtained in the pitch sequence task are pre-sented as a function of sequence length for amusicand control groups. Vertical bars indicate standarderror.

memory, the performance obtained for the 2-and 3-tone distance with interpolated tones wasassessed. The effect of interference was largerin amusics (Fig. 1A) than in controls (Fig. 1B),especially for a pitch distance of 2 tones (t16 =–2.95; P < 0.01). The overall interaction be-tween group and pitch distance was significant[F (2, 32) = 4.08; P < 0.05].

In the pitch sequence task, amusics scoredbelow controls irrespectively of sequence length[F (1, 12) = 11.43; P< 0.01; see Fig. 2]; however,the interaction between group and sequencelength was not significant [F (2, 24) = 2.22;P > 0.13].

Finally, the scores obtained in the interfer-ence condition of the single-tone comparisontask (averaged over the pitch distance of 2 and3 tones) were correlated with the scores ob-tained in the pitch sequence task (averaged oversequences of 1, 3 and 5 tones, with r12 = 0.80;P < 0.005 as computed over both groups). Thesame correlation computed in the amusic grouponly tended to reach significance (r5 = 0.74;P < 0.061).

Conclusion

Amusics’ performance on both pitch mem-ory tasks was impaired relative to controls.Part of this memory impairment might be dueto a perceptual pitch deficit because amusicsperformed below controls in the least demand-

ing memory comparisons (those involving sin-gle tones). However, this pitch deficit seems am-plified by memory load and interference. Thus,the results confirm the presence of a pitch mem-ory problem in congenital amusia. This pitchmemory difficulty is characterized not only bya difficulty in retaining pitch over time, as pre-viously suggested,5 but also by a greater suscep-tibility to interference.

These results are also consistent with struc-tural imaging data showing that amusic indi-viduals have a reduction in white matter andenhancement of gray matter in the right frontalinferior gyrus,7,8 in a region that has been previ-ously associated with pitch memory in a single-tone pitch comparison task.9

Conflicts of Interest

The authors declare no conflicts of interest.

References

1. Peretz, I. 2008. Musical disorders: from behavior togenes. Curr. Dir. Psychol. Sci. 17: 329–333.

2. Foxton, J.M., J.L. Dean, et al. 2004. Characterization ofdeficits in pitch perception underlying tone-deafness.Brain 127: 801–810.

3. Hyde, K. L. & I. Peretz. 2004. Brains that are out oftune but in time. Psychol. Sci. 15: 356–360.

4. Peretz, I., E. Brattico & M. Tervaniemi. 2005. Abnor-mal electrical brain responses to pitch in congenitalamusia. Ann. Neurol. 58: 478–482.

5. Griffiths, T.D., C. McDonald, et al. 2007. Could acongenital disorder of musical perception ever be ex-plained by a single gene? Presented at the Interna-tional Workshop on the Biology and Genetics of Mu-sic. Bologna, Italy.

6. Peretz, I., A.S. Champod & K. Hyde. 2003. Vari-eties of musical disorders: the Montreal Battery ofEvaluation of Amusia. Ann. N. Y. Acad. Sci. 999: 58–75.

7. Hyde, K.L., R.J. Zatorre, et al. 2006. Morphometry ofthe amusic brain: a two-site study. Brain 129: 2562–2570.

8. Hyde, K.L., J.P. Lerch, et al. 2007. Cortical thicknessin congenital amusia: when less is better than more. J.

Neurosci. 27: 13028–13032.9. Zatorre, R.J., A.C. Evans & E. Meyer. 1994. Neu-

ral mechanisms underlying melodic perception andmemory for pitch. J. Neurosci. 14: 1908–1919.